With rapid deployments of photovoltaic (PV) systems, imbalances between energy supply and demand become increasingly pronounced. Air conditioning is a major consumer of electricity and a key energy flexible resource to improve PV onsite consumption. This study developed and evaluated four easy-to-deploy indoor temperature reset strategies for air-conditioning systems, including a time-of-use strategy and three adaptive strategies, based on a building simulation platform of a typical office building in Guangzhou. Results showed that the adaptive indoor temperature setpoint reset strategies effectively alleviated mismatches between PV power and electric load of air-conditioners. The adaptive strategies increased PV self-consumption and self-sufficiency by 5.4

Accurate prediction of operative temperature, which is a combined measure of air temperature and the average temperature of surrounding surfaces as felt by occupants, is crucial for optimizing the performance of heating, ventilation, and air-conditioning (HVAC) systems, improving occupant comfort, and reducing energy consumption, particularly in sentry buildings. This study proposes a hybrid deep learning framework, termed the Hybrid LSTM-Transformer, which combines Long Short-Term Memory (LSTM) and Transformer architectures with a Temporal Attention Pooling mechanism. The LSTM component captures short-term dependencies and nonlinear dynamics, while the Transformer models variable interactions and long-term temporal dependencies. Temporal Attention Pooling further highlights critical time steps to extract the most relevant temporal features. High-resolution data were collected from a wall-mounted electric radiant heating system. Comparative experiments demonstrate that the proposed Hybrid LSTM-Transformer for indoor operative temperature forecasting achieves superior performance and stronger robustness against error accumulation compared with state-of-theart algorithms, including LSTM, Artificial Neural Network (ANN), Decision Trees (DT), Extreme Gradient Boosting (XGBoost), and Random Forests (RF). The model achieved an R2 value above 0.87 for operative temperature prediction at a 60-minute horizon. These findings offer valuable insights into the feasibility of developing accurate models for predicting indoor operative temperatures.

The electrification of space heating through heat pump (HP) technologies plays a pivotal role in achieving low-emission and energy-efficient buildings. However, the resulting increase in electricity demand introduces challenges for grid stability, particularly during peak demand periods. Integrating thermal energy storage (TES) systems into HP configurations offers a promising solution to address these challenges by enhancing both energy efficiency and demand-side flexibility. This review study systematically investigates the integration of TES units into HP systems specifically for space heating applications, with a particular focus on energy efficiency improvement and strategies for demand management such as peak clipping and load shifting. The study discusses the concept of energy flexibility and its relevance to heating systems. The core of the review categorizes TES-HP integration based on the placement of the TES unit: either on the evaporator side or condenser side of the heat pump. Literature findings reveal that TES integration at the evaporator side, particularly using phase change materials (PCMs), significantly improves system efficiency and reduces compressor runtime. Conversely, TES at the condenser side is more effective in achieving load shifting and demand response targets. Key insights include the importance of PCM selection, TES configuration, and advanced control algorithms. While current applications show great promise, future developments are expected to focus on smart control integration, improved PCM thermal properties, and cost-effective implementation to optimize both performance and grid interaction.

Dental clinics are enclosed environments where patients generate significant quantities of aerosolized pathogens during procedures. Under low-velocity ventilation conditions, indoor airflow is strongly influenced by human thermal plumes, which hinder effective pollutant removal. This study combines numerical simulations with experimental validation to analyze the coupled effects of wall-attached jets and human thermal plumes on airflow and aerosol transport in a single-chair dental clinic. The air supply is provided through linear slot diffusers mounted on the ceiling with dimensions of 1.5 m & times; 0.05 m. The results show that, at low supply velocities, the ventilation jet and thermal plume interact, with optimal coupling occurring at a supply air velocity of 0.6 m/ s. Beyond this velocity, plume integrity deteriorates due to jet-induced disruption. Mechanistic analysis identifies the transition from buoyancy-dominated to jet-dominated flow between 0.6 and 0.8 m/s. At 0.6 m/s, a dynamic balance is achieved between jet momentum and thermal buoyancy, producing a maximum plume rise velocity of 0.24 m/s. When the supply air velocity exceeds 0.8 m/s, increased recirculation and plume fragmentation reduce the effectiveness of directional aerosol transport. To address higher ventilation requirements, the study further optimized vent configurations. One scenario was identified in which 68.75% of the clinician's breathing-height plane exhibits aerosol concentrations below 27% of the initial value, making it the most effective solution. The research results provide reference for optimizing ventilation design in dental clinics and propose practical strategies for controlling aerosol transmission pathways and reducing the risk of cross-infection.

The intelligent operation of building energy systems (BES) is crucial for energy sustainability, but is increasingly vulnerable to disruptive cyber-attacks, such as false data injection. Existing control strategies lack the adaptability to maintain efficient and secure operation when critical sensor measurements are compromised. Therefore, this work proposed a physics-aware embedded adaptive model predictive control (MPC) framework guided by reinforcement learning (RL) to improve operational cyber-resilience of BES. Firstly, a data-driven Kalman detection filter was designed to identify and mitigate indoor temperature measurement anomalies caused by false data injection attacks, which can be integrated into the control optimization and provide reliable indoor temperature estimates to downstream controllers. Secondly, an RL-guided adaptive MPC algorithm was developed to solve the optimal energy management problem, where the quantile regression model based on the reference trajectory from the RL agent and historical decisions was employed to dynamically narrow the MPC’s feasible boundaries of control variables and accelerate the optimization process. Thirdly, a physics-aware sparse neural network-based thermal dynamic model was presented to efficiently capture building thermal responses. The trained model was then transformed into a mixed-integer linear programming formulation, allowing seamless incorporation into the optimization framework and efficient real-time prediction. A case study building was conducted to demonstrate the effectiveness of the developed framework. The results showed that the proposed method is capable of mitigating the impact of cyber-attacks on the energy management of BES and achieving the improvement of computational efficiency. The total operating costs and energy consumption by using the proposed method were reduced by 46.7% and 11.5%, on average, compared to other baseline control approaches under cyber-attacks. These results demonstrate that the proposed approach provides a robust and computationally efficient solution for secure and optimal energy management in cyber-physical building environments.

Integrating hybrid renewables, stationary battery storage, and electric vehicles is critical to enhancing the flexibility and sustainability of future energy systems. However, uncertainties in renewable generation, energy demand, and electric vehicle behaviors, coupled with complex grid interactions, pose significant challenges for system design and operation. To address these issues, this study developed an uncertainty-aware optimization framework that embedded grid-supportive control into the design process, incorporated an incentive-compatible vehicle-to-building compensation mechanism reflecting electric vehicle battery degradation, and introduced a grid-friendly flexibility indicator to assess temporal grid interactions. Multi-objective optimization under stochastic scenarios was conducted to identify Pareto-optimal configurations, followed by entropy-weighted multicriteria decision-making for final selection. Compared with the conventional strategy without grid-supportive operation, the proposed framework achieved a 26.6 % reduction in CO2 emissions and 47.3 % decrease in operational costs. It also improved the load match ratio and grid-friendly flexibility by 8.5 % and 122.9 %, respectively, with only a 6.8 % decrease in renewable utilization. Additional simulations verified the reliability and effectiveness of the proposed framework under uncertainties, yielding annual averages of 34.8 % gridfriendly flexibility, 84.7 % renewable utilization, 87.6 % load-match ratio, and 6,822.9 kg CO2 reduction across all stochastic scenario sets. This study advances beyond existing studies by coupling uncertainty-aware design optimization with a grid-supportive control strategy, offering a practical and sustainable pathway for deploying gird-friendly and reliable integrated energy systems with renewables, storage, and electric vehicles in real-world applications.

This study aims to characterize and optimize the thermal comfort, environmental impact, economic performance, and energy flexibility of a thermally activated building system (TABS) integrated net/nearly zero energy building (NZEB) equipped with ground source heat pump-air source heat pump-photovoltaic-battery (GSHP-ASHP-PV-battery) systems. The TABS parameters, including design variables, physical properties, and operating variables, were considered in the performance investigation. The developed TABS model was validated and exhibited acceptable performance based on two in-situ response experiments. The Taguchi method and Utility concept were used to design simulation cases and determine the optimal conditions. The simulation results from the Taguchi confirmation tests demonstrated improvements across all performance indicators in both heating and cooling operations through appropriate optimization. In particular, the Taguchi-optimal cases achieved reductions of up to 58.3% in the heating unmet-hour ratio and 28.4% in the heating imported electricity cost, compared with the best-performing cases in the design set. The performance of each Utility-optimal case was comparable to, or slightly lower than, that of the Taguchi-optimal cases. Nevertheless, the Utility-optimal cases for the unmet hour ratio, carbon dioxide emissions, imported electricity bill, and self-consumption ratio still outperformed the best-performing cases among the original Taguchi cases, with improvements ranging from 1.5% to 18.3% for cooling and from 9.0% to 38.9% for heating. Additionally, based on the Utility concept, the operating factors exhibited significant influence on various performance indicators, particularly the system start time and operating duration. The results can provide valuable information to facilitate the efficient design and operation of the integrated building energy systems.

Building energy consumption accounts for up to 40% of global energy use, contributing substantially to greenhouse gas emissions. Although demand-side management (DSM) and intelligent control systems have been extensively studied for residential buildings, smart buildings still face critical challenges due to their complex energy consumption profiles and lack of flexible, multi-energy management frameworks. This paper presents a novel multi-objective coordinated energy management system (EMS) that integrates a preference-aware reinforcement-learning framework to balance energy cost and indoor thermal comfort by dynamically coordinating multiple building systems, including heating, ventilation, and air conditioning (HVAC), energy storage systems (ESS), and hydrogen storage tanks (HST). To support day-ahead operation under uncertain renewable generation, the proposed EMS incorporates an AI-based extreme learning machine (ELM) model enhanced by a historical data correlation evaluation (HDCE) method, which improves the accuracy of short-term photovoltaic (PV) generation predictions. Simulation results show that the proposed framework can achieve approximately 13.4% energy cost reduction while maintaining nearly negligible user discomfort. In addition, the integrated PV forecasting module achieves a prediction error of 3.33%, which provides reliable renewable input for the coordinated EMS.

Hydrogen is a clean and sustainable energy carrier with significant potential to reduce fossil fuel dependence and mitigate energy shortages. This study proposes a novel off-grid integrated energy system (IES) for remote cold regions, incorporating solar-driven water electrolysis, hydrogen fuel cell power generation, and hydrogenenriched methane combustion. A dynamic model was developed to evaluate system performance for electricity, heating, and gas supply. A multi-objective optimization framework was introduced, incorporating equal weight and entropy weight-TOPSIS methods to determine the system sizing. Under the two schemes, methane consumption was reduced by 18.8 % and 13.6 %, respectively. The primary investment difference was the hydrogen storage tank size, 1200 m3 for equal weight and 1300 m3 for EWM-TOPSIS, resulting in a 5.36 % higher initial cost for the latter. Following optimization, the ideal sizes for photovoltaic panels, electrolyzer, gas tank, battery, and fuel cell were identified. Comprehensive static economic and annual energy flow analyses confirm the system maintained indoor temperatures around 20 degrees C during the heating season, utilizing solargenerated hydrogen and low-emission hybrid combustion. The proposed solar-hydrogen-electricity-thermalbased IES provides a feasible and efficient pathway for clean energy utilization in off-grid cold regions and supports the broader deployment of hydrogen-based technologies.

District heating (DH) systems play a crucial role in delivering efficient and sustainable thermal energy. The integration of phase change material (PCM) thermal energy storage can enhance their operational flexibility. However, effective control of such systems remains challenging under fluctuating spot price conditions due to the complex interplay between storage dynamics, price variability, and operational constraints. This study proposes a novel method by integrating deep reinforcement learning (DRL) and parametric rule-based control (DRL-RBC) strategy for thermal storage management in DH systems. A grey-box surrogate model is developed to alleviate the dependence on large datasets for DRL training by integrating physical knowledge with validation using small datasets. To enhance adaptability under fluctuating spot price conditions, the proposed strategy combines a DRL agent with a parametric rule-based strategy, where the DRL agent dynamically optimizes the thresholds of the rules to enable adaptive charging and discharging of PCM storage. Unlike most existing studies, this study first trains the DRL agent using the surrogate model and subsequently transfers it to TRNSYS-Python co-simulation for evaluation. Based on a case study, it is shown that, over a one-month evaluation, total heating costs decreased by 1.84%, with weekly savings up to 3.66% compared with daily time-based control. While standalone DRL achieved higher short-term savings of 4.00% in the first week, it exhibited lower stability over a month, confirming the superior robustness and adaptability of the proposed hybrid approach. The proposed strategy offers a pathway for advanced DH system control with PCM thermal energy storage, bridging simulation-based research and real-world application.

To elucidate the evolutionary characteristics of the current available capacity of lithium-ion batteries, this study proposes an online identification framework of characteristic change points (CCPs) for capacity degradation curves. Firstly, health indicators are extracted from a short period of data before the end of constant current charging. The interaction of three data-driven algorithms and multiple feature intervals in terms of capacity estimation accuracy is systematically investigated. Then, the parameter combinations for the improved cumulative sum (CUSUM) algorithm are selected and validated through single-factor analysis, orthogonal experimental design, and main effect analysis. Specifically, the improved CUSUM algorithm exhibits an average error of only 16.42 cycles and an average latency of merely 4.20 cycles. Finally, the proposed improved CUSUM algorithm is applied to identify the CCPs in the battery capacity degradation curves of different morphological types. The validation performance of the proposed algorithm demonstrates its capability of identifying both routine degradation patterns and anomalous transitions. The identified CCPs can serve as evaluation indicators for multiple application scenarios, such as battery design, usage strategy optimization, and second-life utilization.

In ionic liquid hydrogen compressors, the liquid provides functions of sealing, piston lubrication, and hydrogen cooling. However, some liquids are expelled with hydrogen during the discharge, leading to a gradual decrease in liquid volume. Liquid replenishment is therefore necessary to maintain an adequate liquid level for effective sealing and lubrication, while also enhancing cooling efficiency and assisting the compression process closer to isothermal conditions. In this study, a transient two-phase flow model was developed to evaluate the thermal performance of the compressor under three different liquid replenishment schemes. The volume of fluid method was employed to track the interface between hydrogen and ionic liquids. Meanwhile, the two-phase flow and heat transfer were simulated over a complete compression cycle. Results showed that replenishment improved compression performance, achieving a minimum polytropic index of 1.145 and a maximum isothermal efficiency of 86.06%. However, replenishment occupied part of the suction stroke and increased liquid discharge, reducing volumetric efficiency, with the lowest value of 69.23%. Among the three schemes, the cylinder-top replenishment method demonstrated an outstanding trade-off between isothermal efficiency (83.51%) and volumetric efficiency (89.92%), making it a promising strategy for practical hydrogen refuelling applications.

Photovoltaic/thermal (PV/T) technology integrates photovoltaic (PV) and photothermal (PT) conversion, enabling the simultaneous generation of electricity and heat while mitigating PV efficiency loss caused by temperature rise. Spectral-splitting based PV/T system could thermally decouple the photovoltaic working temperature and collecting temperature to achieve the higher solar utilization efficiency.In this work, Fe3O4 ferromagnetic nanofluid as the spectral-splitting working liquid was prepared by a two-step method. Increasing particle size and concentration significantly enhanced the photothermal conversion efficiency but reduced the photovoltaic conversion efficiency. Notably, ferroparticles would alter their distribution behavior under external magnetic field (MF), thus the smart spectral-splitting working liquid enable dynamic tuning of electrical/thermal output by modulating the intensity and direction of MF. The results indicated the PV and PT efficiencies could simultaneously improve though under MF. As the MF intensity increased, the thermal-to-electrical ratio of PV/T system decreased. This is more sensitive response for the horizontal MF due to the nanoparticle chains formed a continuous light absorbing layer while also increasing incident light transmittance. As a result, the overall solar energy utilization efficiency reached its peak of ∼ 92 % at MF intensity of 75 mT when the size of the Fe3O4 nanoparticles was 20 nm and the concentration of ferrofluid was 50 ppm. Moreover, the thermal-to-electrical ratio varying from 5 to 15 as increasing horizontal MF intensity and from 5.4 to 15 as vertical to horizontal MF, implying the magnetic field enables effective dynamic turning the electrical/thermal output of the spectral-splitting PV/T system.

Adsorption thermal energy storage (ATES) is one of the most important ways to realize the efficient utilization of solar energy. The adsorption reaction wave model revealed the heat and mass transfer process in the ATES reactor. However, the existing adsorption reaction wave model can only be used to calculate the overall performance of the ATES reactor with a stable output temperature because the reaction wave mechanism is not yet clear. In this paper, the adsorption rate wave transfer equation was developed using the mechanical wave transfer equation as a reference, which could be used to predict the thermal characteristics of reactor, including the overall performance and physical parameter distribution within the reactor. The results indicated that the maximum prediction deviation of the ATES reactor overall performance was only 6.1 % compared with experimental measurement. The evolution of moisture concentration and temperature in the reactor was characterized, and the minimum coefficient of determination of the predicted physical parameter distributions within the reactor reached 0.967. This prediction method bridged the gap that existing reactor performance prediction methods limited to stable output temperatures, while also predicting detailed physical parameter distributions in the reactor.

Integrating electric vehicles (EVs) into net/nearly zero energy buildings (NZEBs) is crucial for fostering sustainable and resilient energy systems. However, this integration poses challenges due to stochastic EV usage patterns, uncertain renewable generation, and fluctuating building demand, complicating energy coordination among NZEBs, EVs, and the grid. Vehicle-to-building (V2B) technology enhances NZEB performance by enabling bidirectional energy flows, yet battery degradation from V2B and the need for appropriate compensation are often overlooked. This study proposes a flexible energy framework that integrates renewable energy, electricity storage, and EVs, while maintaining grid-friendly interactions. Additionally, a dynamic V2B pricing model is developed considering battery degradation. The simulation results demonstrated that the proposed framework could foster mutually beneficial outcomes for both NZEBs and EV owners. Its reliability and robustness were validated under various uncertainties, achieving an annual average operational profit of 1332.3 USD (0.0067 USD/kWh), a load match ratio of 80.1 %, and a carbon emission reduction of 49,431.8 kg (0.25 kg/kWh), which positively affected the operation of power grids. By facilitating the reliable and cost-effective integration of V2B technology into NZEBs under uncertainties, this study provides valuable insights into fostering an interconnected, intelligent, and resilient energy ecosystem. The findings offer practical guidance for scaling up sustainable energy trading within zero energy communities and cities, contributing to ongoing efforts toward the development of sustainable and low-carbon building energy futures.

Hydrogen compression is crucial in the entire hydrogen economic chain. The ionic liquid compression system has demonstrated advantages in the hydrogen pressurisation, where the rational design of the hydraulic-drive subsystem is the key to achieving efficient gas compression functionality. However, a relatively high loss is observed due to the overflow discharge of the hydraulic oil. This paper proposes a new compression system that integrates the hydraulic accumulator for energy-saving. Through numerical simulations, the study first investigates the influence of the oil compressibility on the system design. The findings reveal that hydraulic oil compressibility must be factored into the compression system design, as this parameter induces an 8.22% discrepancy in mass flow rate measurements. Then, the optimisation of design parameters of the hybrid compression system is carried out. Under the optimal combination of 3 MPa precharge pressure, 2 pressure ratio, and 0.2 L total volume, the system achieves a highest exergy efficiency of 73.33%. Based on the optimal parameters identified, this paper further explores the impact of design variables on system efficiency, with the objective of elucidating the underlying mechanisms through which individual parameters govern performance outcomes. Furthermore, the exergy analysis is carried out, which reveals that directional valve losses constitute the main source of exergy losses in the system, accounting for 66.71% of the total exergy losses, indicating the primary optimisation direction for further energy saving. The study provides insights and inspiration for designing an energy-saving and economic hydrogen compression system.

A data-driven model predictive control (MPC) strategy embedded with rule mining was proposed to discover optimal relationships between Indoor Air Quality (IAQ) events and operations of Heating, Ventilation and Air Conditioning (HVAC) systems to optimize IAQ, building thermal comfort, and energy consumption. In this strategy, a rule mining method was used to discover the relationships between occurrences of IAQ events and optimal HVAC operations, including occurrence time rules, co-occurrence rules and sequential occurrence rules. An encoder-decoder Long Short-Term Memory (LSTM) model was used to predict future building performance, and an event detection method was developed to identify the occurrence of pollutants’ events based on the prediction and real-time observations. With the detected occurrences of events, the rules derived from the rule mining method were used to provide preconditioned fuzzy optimal HVAC operations, which were then used to improve the Firefly algorithm (FA) to generate control settings. Simulation tests based on a house with a cooling system showed that, by using the MPC strategy, the pollutants’ peak concentrations of CO2, NO2 and PM2.5 were reduced by 25.4 %, 22.8 % and 35.3 %, respectively, compared with those using the baseline strategy. The exposure times of high concentrations of CO2, NO2 and PM2.5 were reduced by 400 min, 50 min and 55 min and 8.8 % energy savings were achieved. The HVAC energy consumption using MPC with rules was 5.1 % lower, and the pollutants’ peak concentrations of CO2, NO2 and PM2.5 were 13.2 %, 23.4 % and 22.3 % lower, respectively, in comparison with using MPC without rules.

As both buildings and power systems undergo rapid decarbonization, demand flexibility (DF) in buildings has emerged as a key enabler for renewable energy integration, grid reliability, energy resilience, and efficient and low carbon operation. By dynamically adjusting building energy use in response to grid requirements, DF can help reduce building operational costs, respond to renewable variability, reduce peak demand, alleviate network congestion, and enhance overall system stability. Consequently, significant research efforts have examined building DF from multiple perspectives, including flexibility definitions, resources, characterization, assessment, optimization, and application scenarios. However, the effective delivery of DF in buildings requires an integrated, end-to-end approach spanning perception, cognition, decision, execution, and verification, which remains a critical gap in the existing literature. This review addresses this gap by proposing a layered architecture that integrates these stages into a coherent framework and provides insights from existing studies in this field. It synthesizes enabling methods, clarifies cross-layer interactions, and examines how data, models, control strategies, implementation mechanisms, and evaluation approaches collectively support DF delivery. Critical cross-layer challenges are identified, and future research priorities for robust and trustworthy DF delivery are outlined. This work establishes a conceptual foundation for advancing interoperable, grid-responsive, and practically deployable DF solutions in buildings, supporting both building decarbonization and more efficient, resilient grid operation

Airflow management in air-cooled data centers is commonly improved through room-scale supply-air optimization, aisle containment, floor-tile adjustment, and rack-level layout design. However, comparatively less attention has been paid to the initial direction and momentum characteristics of server exhaust airflow, which can strongly influence hot-air recirculation, thermal stratification, and rack-level hotspot formation. This study investigates tilted server arrangement as a source-side airflow organization strategy for improving thermal-hydraulic performance in data centers. A refined room–rack–server computational fluid dynamics model, validated using cabinet-scale experimental measurements, was developed to evaluate airflow and temperature distributions under server tilt angles of 0°, 15°, 30°, and 45°. Standard thermal performance indicators, including the Return Temperature Index and Index of Mixing, were combined with server-scale Nusselt number and room-scale Archimedes number analyses to clarify the competing effects of server inclination on internal convective heat transfer and room-level exhaust transport. The results show that server tilting improves the directionality of exhaust airflow and suppresses hot-air accumulation near the cold aisle. Among the investigated cases, the 30° tilted arrangement provides the best overall performance, increasing the Return Temperature Index from 0.781 to 0.914 and reducing the Index of Mixing from 0.376 to 0.030 compared with the horizontal configuration. Although increasing the tilt angle slightly weakens internal convective heat transfer, an appropriate tilt reduces the Archimedes number of the exhaust plume and promotes inertia-dominated jet transport toward the return outlet. The increased return-air temperature indicates a potential cooling-energy saving of approximately 4.0%-10.0%, based on previously reported empirical relationships. These findings provide a useful reference for improving rack-level thermal management and developing source-side airflow organization strategies in air-cooled data centers.